When investigating natural language processing, we must first outline the various difficulties attributed to writing programs that "understand" natural language. We note the ambiguity, imprecision, and inherent informality of natural language, and the gap between that ambiguity and the precision and formality of languages ​​used in software development for analysis, design, and programming. Bridging the gap between informal natural language and formal conceptual models is critical with applications designed to “understand” natural language. Secondly, some of the applications developed over the decades to “understand” natural language will be described. We will look at the groundbreaking work of the 1960s with ELIZA, which used a rudimentary parser to imitate a therapist until the 1980s, and the great strides made by IBM in language translation of interfaces until the birth of WATSON, the superhuman. interface program that beat two expert human players thanks to its huge knowledge bank and SIRI with its voice recognition capabilities. Finally, we will look at the various techniques and theories used in natural language processing to create applications that "understand" human language. Namely, theories of parsing, semantics, word sense disambiguation, corpus linguistics, and lexicon. Barriers to Understanding Natural Language in Applications In any interaction with a computer, we use natural language to describe our needs and problems. The natural language we use to describe our problems and requests is often complex, vague and ambiguous. Human sentences are complicated and complex. This is especially true when they contain clauses and sentences that describe and relate different objects, conditions, and actions. Sen...... middle of the paper...... its structure from its constituent parts. We have emphasized that ambiguity in text is one of the main problems encountered with NLP ambiguity in natural language, and that word sense disambiguation is the theory developed and implemented to counter this main barrier. Finally we studied semantics, or the study of the meaning behind words. We've noticed that understanding natural languages ​​is just as important as syntax. We have noticed that any application needs semantic theory to guide understanding and create an efficient human interface. In investigating the various barriers associated with natural language understanding and applications in the discipline over the decades and ultimately some of the techniques or methodologies implemented by these applications to successfully understand natural language, we have been able to investigate in-depth and complete on natural language understanding..